Th e Argument from Consciousness Revisitedscotus/files/O'ConnorKimbleOSPRv.III.pdf7 Th e Argument...

7 Th e Argument from

Consciousness Revisited

Kevin Kimble and Timothy O’Connor

More than two decades ago, Richard Swinburne and Robert Adams put forth an argument for theism that they aptly labeled the argument from consciousness . 1 A thumbnail sketch of the argument goes like this: Th ere are facts involving correlations between brain states and conscious states of per-sons for which rational inquiry demands a satisfying explanation. Th ere are but two broad forms such a possible explanation may take: the correlations can be explained either through more basic scientifi c laws or by the inten-tions and actions of a powerful personal agent. Since the correlations’ facts cannot be given an adequate scientifi c explanation, the best explanation is that they are the result of the work of a purposeful agent.

Our aim in what follows is twofold. First, we consider sophisticated recent attempts in the philosophy of mind to defend a robustly physicalist account of the phenomenal character of experience, accounts that, if suc-cessful, would undercut the core premise of the argument from conscious-ness (AC). We will try to show, however, that these accounts fail. We then consider the version of AC advanced by Swinburne and Adams. We con-tend that their versions are defective, since they overlook a naturalistic form of explanation that is available even on a robustly dualistic view of con-scious states. However, we go on to show that the argument may more plausibly be recast by treating the very form of explanation of conscious states we outline as a further datum in the currently popular fi ne-tuning version of the design argument. We do not attempt to determine whether the fi ne-tuning argument is ultimately successful.

1 Swinburne ( 1979 , 2004 ) and Adam ( 1987 ) . Th eir arguments develop a line of thought presented in cruder fashion by John Locke in the fourth book of An Essay Concerning Human Understanding ( 1690 ) .

Th e Argument from Consciousness Revisited 111

I. DIRECT PHENOMENAL AWARENESS

Th e fact that there are strong correlations between underlying neuro-physical states of the brain and conscious mental states is not in serious doubt. Philosophers of mind commonly divide the latter into two broad classes, inten-tional states and phenomenal states. Th e former has commonly been equated with propositional attitude states (beliefs, desires, and so forth), while the latter has been thought to consist of the “raw feels” of experience—the qualitative way things look, sound, taste, and so on. Although we seriously question this bifurcation of mental states into these two distinct kinds, for our present pur-poses we may safely focus solely on the feature of phenomenal character. 2 Although the present state of neuroscience does not yield precise hypotheses, it is highly plausible that certain neuro-physiological types directly correlate with certain phenomenal types. So we may suppose that brain state R correlates with a phenomenally reddish look, brain state B correlates with awareness of phenomenal blue; T correlates with tangy orange sensations, P with a certain piercing auditory quality, and so on. While these cartoonish schema certainly oversimplify how mind-brain correlations actually work, and the way in which brain states and phenomenal states map onto each other, we can assume some-thing like this for the sake of simplicity and concreteness.

Physical-phenomenal correlations of this kind can count as evidence for design only given an assumption that phenomenal properties are ontologi-cally distinct from their physical-functional correlates. Unless the most basic phenomenal properties are ontologically novel features of the world, funda-mental features over and above the features characterized by physical theory, the argument that the correlations constitute a credible design-type fact to be included among other design evidence will be open to a straightforward refu-tation. Identity is the most satisfying explanation of the correlations that is imaginable, and hence one that cannot be improved in theological terms.

Th us the proponent of the argument from consciousness needs to make plausible that phenomenal qualities themselves are ontologically primitive. Swinburne and Adams devote little space to this question, and the discus-sion they do carry on is dated. In particular, there are two prominent mate-rialist accounts of phenomenal qualities in recent philosophy of mind, unanticipated by Swinburne and Adams, that we will consider here.

First, some contemporary thinkers insist that the connection between certain physical and phenomenal properties is one of identity, despite the

2 For a recent challenge to the now-orthodox bifurcation, see Horgan and Tienson ( 2002 ) and Kimble ( 2006 ) .

112 Kevin Kimble and Timothy O’Connor

fact that we cannot grasp the connection a priori. Th ey point out that we may conceive of a single property in more than one way, corresponding to diff erent modes of epistemic access to it. In the case of physical/phenomenal features, we have two wholly distinct sets of concepts corresponding to the modes of scientifi c theoretical identifi cation and introspective awareness. Th ese thinkers then argue that, due to the peculiar nature of our phenome-nal concepts , the fact that we cannot discern the physical-phenomenal iden-tity via a priori refl ection is unsurprising, and it does not follow that the properties we variously conceive of are ontologically distinct. 3 We shall call this strategy type physicalism .

Second, a very diff erent strategy for identifying phenomenal qualities with physical properties has developed in recent years, motivated by a view of phenomenal experience called representationalism . Representationalism is the general term for a family of related positions, but the core idea among physicalist versions of the theory is that phenomenal qualities are to be identifi ed with certain externally constituted representational properties of experience, rather than with internal states (physical-functional or primi-tively mental) of the experiencing subject. 4 Phenomenal character is identi-cal to an experiential state’s intentional content which, depending on the particular view, consists of entities such as external properties, proposi-tions, or states of aff airs. On plausible physicalist assumptions, such con-tent involves structured physical properties or dispositions, such as (for colors) the disposition of a surface to refl ect a certain percentage of light from the visible spectrum. Th us, on representationalism, the look of the redness of the rose to a subject is none other than the external physical property of redness. We will call physicalistically acceptable representa-tional properties p-representational properties. Since phenomenal qualities are extra-mental features, there is no primitive mind-body correlation to be explained. 5

Given the growing popularity of these a posteriori physicalist accounts of the phenomenal character of experience, a defender of the argument from

3 See Loar ( 1997 ) , McLaughlin ( 2001 ) , Papineau ( 2002 ) , and Balog (unpublished ms.). 4 See Dretske ( 1995 ) and Tye ( 2000 ) . 5 Higher order varieties of representationalism typically do not attempt a reductive

account of phenomenal character. Higher order theories (e.g., Rosenthal’s “higher order thought,” Papineau’s “higher order perception,” and Lycan’s internal monitoring theories) purport to explain the awareness relation , why a subject is consciously aware of qualia in some cases but not in others. Such theories attempt to explain the nature of phenomenal character itself only insofar as they combine the higher order component with some other theory (e.g. fi rst order representationalism, functionalism, etc.). We do not discuss higher order theories separately here.


consciousness must make the case that all such strategies are untenable. We will now undertake that task.

Th e case for the ontological distinctness of phenomenal qualities from any physical-functional properties is broadly introspective. When I look at a luscious mandarin orange in bright sunlight and bite into it, at the same time focusing my attention on the qualitative way I experience it, I become phenomenally aware of an orange-ish quality presented to me in a sensate visual way—a smooth and uniform, simple feature impressing itself upon and occupying a region of my visual fi eld. In addition, I am aware of a sweet tangy sensation on my tongue, which is, once again, a fairly simple and uniform taste quality. Th ese contents of my awareness are “transparent” to me in the sense that I have unmediated access to what it is like to experience those qualities. And I recognize certain facts about those qualities that bear on the a posteriori identity thesis. To an approximation, let us say that a property is non-structural if and only if its instantiation does not even partly consist in the instantiation of a plurality of more basic properties by either the entity itself or its parts. 6 Now, my phenomenal visual fi eld may be highly complicated, consisting of diff erently sized, shaped, and colored objects arrayed in a certain manner. But I can focus on suffi ciently small subregions of the fi eld that feature nonstructural or simple properties, such as a deter-minate (phenomenal) shade of blue. Quite generally, it seems that the con-tents of phenomenal awareness ultimately decompose into such simple, non-structural features, all of which are themselves objects of awareness.

Our simple argument from direct awareness goes like this: phenomenal awareness reveals an intrinsic simplicity in the elements comprising the quali-tative character of my experience. However, the physical-functional (or p- representational) properties with which that character is supposed to be identifi ed do not possess this intrinsic simplicity. Type-physicalists identify phenomenal character with complex neuro-scientifi c properties, while repre-sentationalists claim that phenomenal qualities are externally constituted low-level physical properties. Both candidates are heterogeneous structural features. Hence, the immediate contents of my phenomenal awareness, the phenome-nal character and qualities of which I am directly aware, are not identical to underlying physical-functional (or p-representational) properties.

Moreover, in my phenomenal awareness of the orange-ish look of the man-darin orange, there must be some explanation for the fact that the phenomenal property I directly grasp is not presented to me either as a neuro-physical state of some sort, or as a surface refl ectance or some other underlying physical

6 Cf. O’Connor ( 2000 : 110) .


property, as a physical color would be in its own nature. Th e advocate of onto-logically primitive qualia has at his disposal a ready explanation for this. Phenomenal properties are to be identifi ed with neither internal physical- functional states of the brain nor extra-mental p-representional properties. By appealing to the notion of immediate awareness or acquaintance , the qualia theorist can make the case that the qualities one has direct access to are intrin-sic, non-structural features of experience. Th is explanation provides adequate grounds for our immediate awareness of phenomenal character and best accounts for its intrinsic simplicity vis-à-vis the objects and content of that awareness.

Now we can deploy this argument from direct access in the task of eff ec-tively debunking recent attempts to defend the a posteriori identity thesis. It is important to notice that the argument we advance here does not make the assumption that every distinct concept expresses a distinct and genuine property, nor does it make use of dubious premises concerning the relation-ship between conceivability and possibility. Hence it is to be sharply distin-guished from other widely known and discussed arguments for property dualism. 7

One physicalist response to our argument contends that what appears to be a phenomenal/physical property distinction is in reality merely a dis-tinctness between modes of presentation of the selfsame property. However, the problem just resurfaces as a question about the nature of the relevant modes of presentation. Suppose that physical blueness is a certain surface refl ectance property B, and that what I am phenomenally aware of—the particular visual qualitative character that is presented to my awareness—is a mode of presentation of B. 8 As the direct object of my awareness, this mode of presentation M is a feature or property of either B itself or one of my representations of B, and thus M stands in need of some acceptable physicalistic explanation or reduction so as to avoid an undesirable “dual-ism” of properties.

Materialists must deny that awareness of phenomenal properties and con-tent involves any distinct mode of presentation of the referent property or content. But how can there be a phenomenal presenting of a property or con-tent to the subject’s awareness unless there is also some mode of presentation

7 Key statements and discussions of arguments for property dualism are found in Smart ( 1959 ) , Jackson ( 1982 ) , Kripke ( 1980 ) , Chalmers ( 1996 ) , White ( 2010 ) , Loar ( 1997 ) , Block and Stalnaker ( 1999 ) , and Gertler ( 2002 ) .

8 For purposes of illustration, our examples of both external world properties and internal neuro-physical properties are used somewhat interchangeably, as the points we make apply equally to both p-representational properties and type-physicalism.


involved? Th e answer, for the a posteriori physicalist, is that the referent property or property-involving content presents itself directly to awareness; the phenomenal property is revealed directly as it is in itself, and the subject grasps this property via the exercise of an appropriate phenomenal concept . It is the peculiar nature of these phenomenal concepts that accounts for why we don’t grasp phenomenal qualities as physical-functional properties. Th ere are various ways this idea has been elaborated.

Th eories of phenomenal concepts fall neatly into two broad classes: caus-al-recognitional accounts and self-referential or quotational accounts. On the fi rst type of account, phenomenal concepts are a special kind of recog-nitional concept that pick out phenomenal qualities directly by demon-strating or naming them, without relying on theoretical or background information. Loar, McLaughlin, and Tye each hold this sort of view. 9 Loar’s and McLaughlin’s proposals are virtually identical, while Tye’s view diff ers from theirs crucially on the issue of the nature of the content of phenome-nal concepts. Loar and McLaughlin hold that the referents of phenomenal concepts are internal neuro-biological properties, whereas Tye holds that their referents are externally individuated representational contents.

Th e self-referential or quotational account is similar to the fi rst type of view except that the phenomenal qualities attended to in experience serve as constituents of the very phenomenal concepts that pick them out, thus con-ferring upon them a self-referring function. Th e basic idea is that there is a subset of neural states that are constitutive of phenomenal states (or serve as the core realizers of such states), and a given phenomenal state may get embedded in a more complex representation by a sort of mental “quotation” process which incorporates the original state as a constituent. Th e representa-tion in turn takes on a sort of refl exive indexical function of referring to the very neural (phenomenal) state which in large measure constitutes the repre-sentation itself. Th e representation is a phenomenal concept. Block, Balog, and Papineau all propose versions of this type of account. While the details of these individual accounts diff er in various places, they are all of one accord in their implicit answer to the problem of intrinsic simplicity. 10

In rough outline, here is how the story goes as concerns our access to and awareness of phenomenal properties. Th e subject of an experience, in

9 Loar ( 1997 ) ; McLaughlin ( 2001 and 2003 ) ; Tye ( 2003 ) . 10 Th e only authors who explicitly address the issue are McLaughlin and Tye, although

some of what others (e.g. Loar 1997 ; Block 2006 ) have to say in the context of addressing the problem of the explanatory gap have a bearing on the issue at hand. We will focus on McLaughlin’s and Tye’s accounts, as well as their related comments.


exercising an appropriate phenomenal concept during an act of introspec-tive awareness, directly grasps the phenomenal property as it is in itself . Type-physicalists such as McLaughlin speak of phenomenal qualities as constituting their own modes of presentation for the subject, thus revealing themselves to the subject’s awareness as they are in and of themselves—yet not by grasping those properties via any essential or non-essential aspects of them. Phenomenal concepts lack any descriptive content whatsoever, func-tioning much like non-descriptive name concepts or type-demonstrative concepts, and hence do not reveal anything about the essential nature of phenomenal properties. 11

Other theorists write as if they deny any mode of presentation at all. Tye, for example, in discussing his view on color, says “Indeed, on my view, colors are not presented to us in sensory experience under any mode of presentation at all. Our awareness is direct.” 12 Th is is true of fi rst-order phenomenal aware-ness as well as introspective awareness of phenomenal content. With respect to the latter, Tye claims that introspecting a perceptual state involves exercis-ing one or more phenomenal concepts that pick out the represented content directly and refer via the direct causal connection they have with their refer-ents. 13 In agreement with McLaughlin, he holds that these concepts have no associated reference-fi xers and contain no descriptive content at all. Rather, “the referent is presented without the assistance of associated features distinct from the referent which the thinker a priori associates with it. Th ere is no separate guise that the referent takes in the thinker’s thoughts.” 14 A subject knows that she is in a visual state “as of” seeing blue just by attending to how things look to her, not by attending to something else connected with it. What the phenomenal concept directly picks out is the complex representa-tional content, the key constituent of which is a surface refl ectance-involving state of aff airs. Th e subject is phenomenally aware of the state of aff airs that has a surface refl ectance b as a component.

Despite some diff erences over how they use terminology, Tye and McLaughlin are in basic agreement on the issue of modes of presentation of

11 McLaughlin ( 2001 : 324–6), ( 2003 : 148) . 12 Tye ( 2002 ) . 13 “In fi rst approximation, a phenomenal concept C refers to a phenomenal quality Q

via C’s being the concept that is exercised in an introspective act of awareness by person P if, and only if, under normal conditions of introspection, Q is tokened in P’s current experience and because Q is tokened” (Tye 2003 ) .

14 Tye ( 2003 ) . Th e examples he uses in this section primarily concern the application of phenomenal concepts to pain, but he explicitly intends his account to generalize to all phenomenal states.


phenomenal properties. When Tye says that a phenomenal property is not presented under any mode of presentation, he means that there is no aspect or property separate from the referent itself that serves to anchor or ground application of the relevant phenomenal concept. Since he apparently accepts the fact that the referent property is “phenomenally presented” to the sub-ject (e.g., there is a bluish look or feel in my visual fi eld), the phenomenal concept picks out the property directly and as it really is, under no other guise. Th is seems to be equivalent to McLaughlin’s claim that phenomenal qualities serve as their own modes of presentation. 15 Of course, as noted previously, the objects of phenomenal awareness are radically diff erent on the two views. McLaughlin holds that phenomenal character is constituted by certain neuro-scientifi c properties of the brain, whereas Tye holds that qualitative character consists of a complex representational state, a crucial component of which are certain low-level physical (surface) properties in the extra-mental environment.

II. AN INCONCLUSIVE ARGUMENT AGAINST A POSTERIORI PHYSICALIST ACCOUNTS OF DIRECT PHENOMENAL AWARENESS

Th e account of direct access to phenomenal properties that we have off ered encompasses claims that stand in tension with the key identity claim made by proponents of a posteriori physicalism. Although we might try to make explicit this tension by formulating these claims in terms of phenomenal concepts, the problem is independent of theories of phenomenal concepts, in that it arises for anyone who accepts a form of “direct access” akin to what we have been calling phenomenal awareness—regardless of whether or not such awareness is mediated by the exercise of phenomenal concepts. Th e following locutions, which we will use somewhat interchangeably, are intended to capture the sense of “direct access” to phenomenal qualities that physicalists such as McLaughlin and Tye allow that we have. Th ey are also intended to be broad enough to permit a form of direct phenomenal access that does not require concept deployment. (If non-conceptual awareness of

15 Block ( 2006 ) also seems to understand these ways of articulating the issue to come down to the same thing. He claims that his theory “involves a notion of a phenomenal concept that has some affi nities with the ‘directness’ story in which there is no metaphysi-cal mode of presentation at all. Th e doctrine that the metaphysical mode of presentation is the same as the referent says there is no metaphysical mode of presentation over and above the referent.”


phenomenal qualities seems inconceivable to you, then no harm will be done by reading the following locutions as claims about our grasp of phe-nomenal properties via the exercise of phenomenal concepts.) When we say that a subject S directly accesses a phenomenal property Q in undergoing a perceptual experience, we mean roughly the following: S is phenomenally aware (“ q-aware ,” as distinct from “ p[perceptually]-aware ”) of Q; or, S phenomenally grasps Q in an unmediated way; or, Q is directly revealed or presented to S in an act of phenomenal or introspective awareness. Th e form of awareness in question might be fi rst-order or higher-order, whereby S directly grasps phenomenal quality Q and conceives it by deploying a phenomenal concept C in an act of introspection.

Now, the following claims made by a posteriori physicalist theories of phenomenal access appear to be in tension:

i. S directly accesses Q as it truly is in itself . Q is revealed to us as it really is, leaving no room for illusion. For example, we are not q -aware of phe-nomenal blue under any distinct mode of presentation or manifesta-tion of it; rather, phenomenal blue is presented to our awareness as it is in itself—meaning that the property (or referent of the appropriate phenomenal concept) itself serves as its own mode of presentation.

ii. Q is a physical-functional property . The object of q -awareness (Q) is an underlying physical property—either a neuro-scientifi c prop-erty within the agent or a micro-physical surface property in the external world. The nature or essence of phenomenal blue consists of either a complex neural state of the visual cortex (type-physical-ism) or a particular surface-refl ectance disposition of an external-facing surface (representationalism).

iii. S does not directly access Q as a physical-functional property . We are not q -aware of phenomenal qualities as physical-functional prop-erties of any sort; they are not presented to awareness as physical-functional properties.

Claims (i) and (iii) seem jointly to preclude claim (ii): if phenomenal qualities are revealed to us in experience as they really are , and if we are not aware of them as underlying physical-functional properties, then how can they be identical with such properties? Alternately, if we directly access phe-nomenal properties as they truly are, and if they are just identical to certain low-level physical-functional properties, then it seems that we should be phenomenally aware of their underlying physical-functional nature.

Th us, there appears to be a straightforward argument against physicalism about phenomenal character. Indeed, this reasoning parallels a similar argu-ment against a posteriori type-physicalism off ered by Horgan and Tienson


( 2001 ) . In circumstances in which a subject S is occurrently aware of a phenomenal quality Q while undergoing a perceptual experience, the argu-ment claims the following:

1. When S is phenomenally aware (Q-aware) of phenomenal quality Q, S is aware of Q as it truly is in itself.

2. When S is q-aware of Q, S is not aware of Q as a physical-func-tional property.

3. If S is aware of Q as it is in itself and S is not aware of Q as a physi-cal-functional property, then Q is not a physical-functional property.

4. Conclusion: Q is not a physical-functional property. 16

Is this apparently straightforward argument against physicalism in fact sound? No; there is a fl aw in the argument that occurs at step (3). Th e con-ditional expressed by (3) is false. It simply does not follow that if S is aware of Q as it is in itself and S is not aware of Q as a physical-functional property, then S is aware of Q as a non -physical-functional property. And this interme-diate step is required if the conditional expressed by (3) is to have any plau-sibility. From the fact that S does not directly access Q as an F, it does not follow that S directly accesses Q as a non-F. In general there is a distinction to be drawn between “ not being aware of a as a B” on the one hand and “being aware of a as a non -B” on the other. A child’s not being aware that the fi gure she is looking at is an octagon does not entail that she is aware of the fi gure as a non-octagon. And her awareness may nevertheless be considered a kind of direct awareness of the fi gure as it is in itself. (A similar point can be made about awareness of certain phenomenal qualities. I may not be aware of a bluish look in my visual fi eld as a property of my experience , but that does not mean that I am aware of the bluish look as being a property of something other than my experience .) In his response to Horgan and Tienson, McLaughlin appeals to a similar distinction, claiming that in order for E to be revealed as an F, the subject must bring E under the concept of F. 17 He points out that we do not grasp phenomenal qualities as physical-functional properties for the simple reason that in grasping the relevant quality via some phenomenal concept C, we do not bring that quality under any physical functional concept. But it does not thereby follow that the phenomenal quality is otherwise than a physical-functional property.

16 Horgan and Tienson ( 2001 : 311) . Although they couch their argument in the lan-guage of phenomenal concepts and conceivability, the essence of the argument could be preserved using the “awareness” locutions we have adopted.

17 McLaughlin ( 2003 : 147) .


In accordance with (ii) above, physicalists maintain that Q is a physical-functional property. Moreover, implicit in (ii) is the idea that the nature or essence of Q is exclusively physical-functional. Denying this would be tanta-mount to conceding that phenomenal properties have dual (or multiple) essences, one of which is non-physical-functional, and this in turn would undercut physicalism in favor of a kind of dual-aspect theory. Th us, physi-calists such as Tye and McLaughlin maintain that the essence of phenome-nal properties is wholly physical-functional.

Tye and McLaughlin have successfully dodged Horgan and Tienson’s argument. Whether their response to it undermines our original argument from direct awareness is something we consider in due course. Let us fi rst observe that we still seek a constructive answer to the following question: given that we directly access a phenomenal quality Q as it really is in itself, and given that the nature of Q is exclusively physical-functional, then why don’t we directly access Q as a physical-functional property? If the direct access we are talking about involves conceptualization of Q, then on one level the answer is that we simply don’t access Q via the exercise of a physi-cal-functional concept. But if Q’s very physical-functional nature is directly revealed to our awareness, then why doesn’t our immediate grasp of Q involve, at least in part, the exercise of an appropriate kind of physical-functional concept to pick it out?

Tye and McLaughlin’s answer is to deny the antecedent in the foregoing question: the essence of a phenomenal property is not revealed to our awareness, and hence we do not directly access that essence. 18 We directly grasp the phenomenal property as it is in itself—the property presents itself as it really is—but this does not mean that we directly access its nature or essence (even in part). And the fact that we don’t directly access phenomenal properties as physical-functional properties implies that we don’t access their nature(s), since they are exclusively physical-functional. Th us, in addition to (i)–(iii), physicalists of the Tye-McLaughlin stripe also hold:

iv. S does not directly access Q’s nature or essence.

Th e way McLaughlin and Tye resolve the apparent tension in our initial claims (i)–(iii) listed above is by off ering a further elaboration on the rela-tionship between (i) and (iii), which involves the additional claim (iv). (iv) qualifi es (i) and helps explain why (iii) is true. Th e fact that we don’t directly

18 Tye ( 2002 , 2003 ) ; McLaughlin ( 2001 : 324) .


grasp the essential nature of phenomenal properties explains why we don’t conceive of them physical-functionally. We may think of (iv) as a denial of the Doctrine of Revelation : 19

Revelation (R) : Th e intrinsic nature of a phenomenal property Q is at least partly revealed to us in our q -awareness of the property.

Th e remaining task for the a posteriori physicalist is to off er a plausible story that explains or clarifi es the consistency between (i) and (iv):

i. S directly accesses Q as it truly is in itself. iv. S does not directly access Q’s nature or essence.

In summary, here is what Tye and McLaughlin say on the matter. Our unmediated grasp of phenomenal properties is attained by exercising non-physical-functional concepts that refer directly to those properties. We are directly aware of Q as it is in itself because our phenomenal grasp of Q involves no distinct mode of presentation of Q, but a direct demonstra-tion of Q itself. Th at is, the phenomenal concept by means of which we access Q demonstrates and presents Q directly. So in presenting itself directly to our awareness, Q serves as its own mode of presentation. Th is is the needed clarifi cation of (i). At the same time, our direct grasp of Q is not by means of some description or mode of presentation of Q that captures or is an aspect of it, or part of its essence, a description or mode that would allow a conceptualization of it as such and such. Our access to Q does not involve fi xing reference via some associated description of Q, since the phenomenal concept that aff ords a direct grasp or access of Q contains no descriptive content. Th e phenomenal concept functions as a naming device or type-demonstrative, and hence does not conceptually reveal anything about the essential nature of Q. So we are not phenome-nally aware of Q as a physical-functional property, since we do not access Q via a physical-functional concept containing the appropriate descrip-tion. Th is provides the required explanation for (iv). Th us according to Tye and McLaughlin, we can be directly aware of phenomenal properties as they truly are in themselves, without illusion, and yet without grasping their natures. Th is explanation renders mutually consistent the central claims (i)–(iv) put forth in their account of direct access to phenomenal character and content.

19 Th e term “revelation” originally comes from Mark Johnston ( 1992 ) , who used it to refer to the thesis that the natures of external color properties are revealed to us in our visual perception of color.


III. THE CONTENT OF PHENOMENAL CONCEPTS

We contend that the Tye-McLaughlin position on phenomenal access and awareness is unstable, in that their position when pressed will either give way to primitivism concerning phenomenal properties, in accordance with the thesis of intrinsic simplicity, or lead to a denial of the phenomenological data as we encounter it in experience. In the end, respecting the phenome-nological evidence requires embracing a limited-revelation thesis, and this picture fi ts best with a primitivist view of the nature of qualitative character.

Let’s consider (i). If we press for greater clarity on issue of what it means for S to directly access Q “truly as it is in itself,” Tye and McLaughlin claim that Q presents itself as it is, as its own mode of presentation. On the face of it, this claim appears to be affi rming a genuine presentation of Q to S’s phenomenal awareness; and since Tye and McLaughlin are phenomenal realists, they take the phenomenology of experience rather seriously. If a phenomenal quality is visually presented to me (e.g., if I directly access it through the visual modality by application of a phenomenal concept appro-priate to that modality), then the quality is presented to me in a certain way, or as something. Th at “certain way” or “as something” is the way it looks to me. True, the mode of presentation just is the referent property itself, but it is nevertheless a substantive presentation of that property. Consider, for example, the experience of staring at the sparkling bluish appearance of the Pacifi c Ocean on a bright afternoon. As one attends to the bluish character of one’s visual fi eld, one’s phenomenal grasp of the bluish quality constitutes a particular way things seem to one, and that way is the direct presenting of the phenomenal property itself, i.e. its constituting its own mode of presen-tation. Th is much is clearly present to one in the experience, and once we have come this far, we can go a step further. Th e presentation of the phe-nomenal quality is something one has a substantive and determinate grasp of, gleaned from the visual phenomenology itself. Th is substantive grasp reveals primitive, non-structural features as the immediate contents of one’s q -awareness. At least that is the way things clearly seem to be, phenomenally speaking. Th e bluish look of the ocean is intrinsically simple in the follow-ing sense: what one directly accesses (what presents itself to one’s awareness) seems to be a smooth, uniform, homogeneous quality, one that is not fur-ther dissectible into parts or discernible patterns, as far as the actual phe-nomenology goes. We have a substantive and determinate conception or grasp of its simple, non-structural qualitative nature. We claim that this feature of intrinsic simplicity thoroughly pervades the phenomenology of


perceptual experience in all of the basic sense modalities. Th is important aspect of the phenomenal character of experience we can call the Intrinsic Simplicity condition (IS):

Intrinsic Simplicity (IS) : Phenomenal character and content is constituted (at least in part) by intrinsic, non-structural features directly accessed in phenomenal awareness.

IS holds for phenomenal qualities grasped via any of the standard sense modalities. Th e sweet, tangy taste of a mandarin orange when you bite into it phenomenally presents itself as a simple, uniform gustatory sensation of a particular kind; a vocal note sung by a choir has a distinctive auditory qual-ity that can be described as a smooth, homogeneous, intrinsic feel imping-ing on one’s awareness. Examples of such intrinsic, qualitative sensations readily come to mind for the other sense modalities as well.

If what we have said so far is on the right track, then in phenomenal and introspective awareness it looks like we do have a positive, substantive grasp of the objects of that awareness, phenomenal properties. Perceptual phe-nomenology reveals an awareness of primitive, non-structural features, in accordance with IS . Th is in turn implies that we have a substantive grasp of the way a phenomenal property really is , which means a grasp of (at least some aspect of ) its real nature or essence. Let’s call this claim minimal or partial descriptivism . (We insert the “at least some aspect” qualifi er in order not to preclude a metaphysical picture on which phenomenal states have dual aspects, primitively phenomenal and physical.)

In introspecting a phenomenal property, either we grasp the property by applying concepts that involve some descriptive content, or we grasp the property via a phenomenal concept, which involves no descriptive content whatsoever. Tye and McLaughlin deny that phenomenal concepts carry any descriptive content or that their application involves conceptualization of the relevant phenomenal quality “as such and such.” Phenomenal concepts in their view are purely recognitional, and so their exercise does not involve the application of other concepts that would allow the subject to conceive of the quality or fi x reference to the quality via a description (including any physical-functional description). But Tye and McLaughlin further assume that any recognitional-phenomenal concept, since it is not descriptive in that way , thus contains no substantive content at all. Here it seems that Tye and McLaughlin present us with a forced choice between two extreme claims: a phenomenal concept either grasps the property it refers to via a full descriptive content that characterizes the property in question or else it grasps the phenomenal property in no substantive way at all. We claim that this amounts to a false dichotomy, because there is a third possibility—there


are recognitional-phenomenal concepts whose exercise does not involve the use of other concepts but which are nonetheless substantive in that they ground the application of further concepts which do partially and accu-rately characterize what the phenomenal property is like.

Th us we contend the phenomenological data suggest that the nature of the content of immediate phenomenal experience lies in between these two implausible extremes—that we possess a robust and determinate grasp of phenomenal qualities (which amounts to more than mere naming or dem-onstration) but at the same time one that falls short of a full description or conceptualization. We do not directly conceive of or access phenomenal properties as instances of full-fl edged, determinate property types of one sort or another (micro-physical, neuro-scientifi c, or otherwise). Nevertheless, we directly grasp them in a substantive way that allows us to form rudimen-tary concepts which truly characterize those properties, and these concepts can be seen on refl ection to be incompatible with their being identifi ed with the sorts of highly structured physical properties or relational functional properties recognized by physics and neuroscience.

To elaborate: the phenomenology of perceptual experience is not com-pletely neutral concerning the nature of the properties that we access in phenomenal awareness. Th e nature of phenomenal experience/the content of phenomenal concepts reveal both the presence of certain features and the absence of other features, which suggests that phenomenal character is not to be identifi ed with underlying physical properties. In the examples con-sidered above, we identifi ed phenomenal color, taste, and sound qualities, all of which exhibit the positive feature of intrinsic simplicity as described in IS . For each of these qualities, we have a robust and determinate concep-tion and grasp of their simple, non-structural qualitative essences. 20 While it is certainly true that we do not conceive of these qualities via any physical-functional description which serves to fi x reference, we nonetheless directly grasp them in such a way that reveals a substantive albeit minimally descrip-tive content grounded in the nature of the phenomenology itself, and which

20 Th e fact that many observers are not cognitively sophisticated enough to have con-cepts such as “simple” and “non-structural” is irrelevant to our argument. What is rele-vant is that so cognizing it would be an appropriate and correct judgment which is directly grounded in the character of the phenomenology. Certainly, one may grasp the “bluish” essence or character of a phenomenal quality while not recognizing it as either simple or non-structural. But in a more sophisticated exercise of one’s introspective abili-ties, one can go on to probe further as to what kind of feature the intrinsic bluishness is. And arguably, it can be discerned as a smooth, homogeneous, non-structural feature. In our awareness of the “feels” of experience, the way things seem is the way things are, because there is no appearance/reality distinction defi ned over such features.


we apply conceptually in introspectively characterizing those features present to our awareness, or how the property is revealed to us. Th is mini-mally descriptive content includes the intrinsic simplicity of the particular phenomenal qualities that we experience.

In being phenomenally and introspectively aware of their qualitative nature, we also may take notice of what is excluded, that is, we notice what is conspicuously absent from the phenomenology of the experience. Th e phenomenology lacks any trace of the heterogeneous structure or disconti-nuity characteristic of any candidate physical states. Take the case of the auditory experience of the musical note. If what we directly (auditorily) grasp is (as Tye contends) an external acoustic wave pattern, then the imme-diate object of our q -awareness is (roughly) a low-level physical state con-sisting of a pattern of vibrations of clusters of air molecules (states of compression and rarefaction) generating a complex waveform comprised of waves with varying wavelength, frequency, pressure, and intensity. But this messy, inhomogeneous, highly composite structure—which is indeed what our experience represents —comes to be completely glossed over in the phe-nomenal character or feel of the experience itself. Similar things can be said of a bluish visual experience of the ocean. If the phenomenal quality we have direct access to is in truth a wavelength refl ectance disposition or pro-fi le involving water’s absorption of longer wavelengths of light and its refl ectance (off a watery surface) of shortwave light energy, then, contrary to what the visual phenomenology suggests, the object we phenomenally grasp is a sharply discontinuous, partite-structured relational property. In both cases, however, the phenomenal feel of our experience strongly points to directly grasped, simple, homogeneous, non-structural features that, in an ordered array, are the immediate objects and content of our awareness. 21

Th e argument remains essentially the same in response to others, such as McLaughlin, who allege that the objects of q -awareness are internal neuro-scientifi c properties. 22 Returning to the example of the musical note, acous-tical information from the waveform is processed by the cochlea and the auditory pathways and structures of the nervous system. Th is is an event involving a highly structured concatenation of electro-chemical impulses and exchanges taking place among billions of neurons in the auditory

21 Note again that we are not denying the eminently plausible claim that phenomenal qualities represent such candidate physical properties in the external world and that these facts constitute our being perceptually aware of such external objects and their properties. Our point is that they cannot be the immediate objects of phenomenal awareness. Th ere is no appearance/reality distinction defi ned over the immediate objects and contents of q -awareness. 22 Lockwood ( 1993 : 274) makes a similar point.


cortex, along with highly complex chemical activity taking place across cell membranes and synaptic connections. Yet this auditory experience has a simple phenomenal structure, given in the musical note itself. At the phenomenal level, the event lacks any remotely recognizable isomorphic structure corresponding to the neuro-physical (let alone micro-physical) com plexity. Th e felt sound quality itself wholly lacks any of the grainy, par-ticulate, discontinuous spatial-temporal structure or composition of its alleged neuro-physiological and ultimately micro-physical correlates.

Th e point is not simply that one’s access to phenomenal qualities glosses over a micro-physical structure which is somehow hidden from one’s aware-ness, but rather that the substantive, positive grasp one does have of them rules out the possibility that the immediate objects of one’s phenomenal awareness have such structure at all. If one wants to claim that our immedi-ate access to phenomenal properties “smooths out” their structure to the point of completely masking their underlying nature, while at the same time granting that the phenomenological data do present us with seemingly prim-itive non-structural properties, then it looks like we are back again to the problem of phenomenal modes of presentation, since there is no appearance/reality distinction defi ned over the immediate objects of phenomenal aware-ness. We grasp phenomenal qualities as they really are, and we grasp them as simple, non-structural properties in the way previously described. Th e phe-nomenology of perceptual experience exhibits the positive feature of intrin-sic simplicity and wholly lacks the structure and particularity that belongs to physical-functional states. Th e presence and absence of these features strongly argues in favor of distinguishing phenomenal character from the underlying physical-functional properties described by physics and neuroscience.

To rebut the point we are making, it is not suffi cient to claim, as Tye and McLaughlin repeatedly do, that the reason we fail to directly access the underlying physical structure of phenomenal qualities is that phenomenal concepts are not physical-functional concepts, and hence we don’t conceive of phenomenal qualities in that way. 23 It is more or less a truism to say that .

23 It is important to be clear about what is and what is not in dispute here. As propo-nents of non-conceptual content, Tye and McLaughlin deny that the very act of grasping a phenomenal property is concept-involving. But Tye and McLaughlin go beyond this rather innocuous claim by further denying that in grasping a phenomenal property (introspectively) via a phenomenal concept, we grasp it in a way that truly captures any part of its nature. Th e only concepts or judgments licensed by the phenomenal concept would be judgments that do not (non-trivially) characterize the nature of the phenome-nal quality itself. (Certain trivial or relational claims about its essence could be allowed, such as “this experience is not bluish like that one,” or “this experience is identical to itself.”) Hence Tye and McLaughlin deny minimal or partial descriptivism.


Tye and McLaughlin claim that in order to be (introspectively) aware of something as an F, you must bring it under the concept of an F. Th at is, they endorse something like the following principle:

Introspection-as (IA) : For subjects S, and objects or properties Q and F, if Q is revealed (in introspection) as an F to S, then S brings Q under the concept of F.

If phenomenal properties are revealed to us in introspection as physical-functional properties (and if we grasp them as such), then we must con-ceive of them via physical-functional concepts. But since, on the phenomenal concept view, we don’t grasp them via physical-functional concepts, it should come as no surprise to discover that phenomenal quali-ties are not revealed as physical-functional properties. Now, we don’t doubt that introspective awareness of and attention to phenomenal qualities requires the deployment of appropriate concepts. Certainly, when you introspect your experience of the ocean and attend to the bluish quality in your visual fi eld, you form and apply an appropriate phenomenal concept that aff ords you a direct and substantive grasp of the blueness you are attending to. In so doing, you need not, and indeed do not, directly grasp the blueness as a physical-functional property. Nor do you directly grasp it as a non -physical-functional property. Instead, you form and apply certain corollary, rudimentary concepts to the property that are directly grounded in what it’s like for you to experience it, immediately “read off ,” as it were, from the very content of the phenomenal concept itself. And that content derives from how the phenomenal quality is directly revealed to your awareness, i.e. as its own mode of presentation. So when you grasp phe-nomenal blue under a phenomenal concept B, then even though B is nei-ther physical-functional nor non -physical-functional, still B is not neutral regarding the nature of the phenomenal property in question. On the basis of B, you can go on to correctly apply concepts such as “is non-structural,” “is a uniform quality,” and so on. Th at is because B itself, and the phenom-enal blueness you directly grasp, has a substantive content that licenses correct application of those corollary concepts. Th is explanation of direct access is consistent with IA ; that is, a non-reductivist or primitivist can grant that Q’s being revealed to S as a non-structural feature requires that S exercise the concept “being a non-structural feature” in grasping it. On this view, we can introspectively grasp the nature of phenomenal qualities in part, and the remaining issue is the question of whether or not that nature is consistent with the property being a physical-functional one.

What we have been arguing is that, with regard to the issue of whether or not our special form of access to phenomenal properties is compatible with their being physical-functional properties, it is not strictly relevant that we


lack direct awareness of a phenomenal quality’s alleged physical-functional structure under some physical-functional conceptualization; rather what matters is that the qualitative content we are positively and robustly aware of gives us a rudimentary grasp of it, which rules out any physical-functional structure as being a component of the quality we access. If the relevant physical-functional properties to be identifi ed with phenomenal qualities are inherently structural and complex, as most theorists think they must be, then the qualities revealed to us in phenomenal and introspective awareness cannot be physical-functional features.

IV. THE ARGUMENT AGAINST A POSTERIORI PHYSICALIST ACCOUNTS OF DIRECT

PHENOMENAL AWARENESS VINDICATED

Th e purpose of the foregoing has been to sketch a plausible position mid-way between the extremes of demanding a full descriptive content for phe-nomenal properties or settling for no substantive content at all. We fi nd the story told here to be an attractive one, cohering well with the phenomenol-ogy of perceptual experience. In fact, as we argue in this section, if we press for further clarity on (iv), insisting that we don’t directly grasp the nature of phenomenal qualities, then in the end we will be forced to deny the phe-nomenological data. Recall that physicalists such as Tye and McLaughlin allege that:

iv. S does not directly access Q’s nature or essence.

Th e nature of phenomenal properties is not in any way revealed to us in phenomenal or introspective awareness. Th e phenomenal concepts we exer-cise in grasping phenomenal qualities have no descriptive content whatso-ever, not even minimally descriptive or substantive content in the way outlined in the previous section. Th e way phenomenal qualities essentially are is not revealed to us, otherwise this would lead to affi rming their intrinsic simplicity. Since they are not revealed to our awareness via any descriptive or substantive mode of presentation, the only thing it can mean to say that phenomenal qualities “are revealed as they really are” and that they “serve as their own modes of presentation” is that they are grasped by direct reference and demonstration. Tye and McLaughlin allude to this function of phenom-enal concepts in several places. 24 Phenomenal concepts “do not conceptually

24 McLaughlin ( 2001 : 324) . See also Tye ( 2003 ) .


reveal anything about the essential nature of phenomenal properties; they simply name or demonstrate them.” So the bluish look of the ocean, how it phenomenally looks to you—call this phenomenal quality b —does not reveal anything of its nature, but is nevertheless revealed directly by your naming or demonstrating b , which itself is a physical-functional property. b is not presented to you in any way or as anything, including either a struc-tural or non-structural property.

Th ere are a variety of ways to think of demonstrative reference. One type of view posits two distinct states: a state C consisting of the exercise of a phenomenal concept, along with a distinct phenomenal quality or state Q that C names or demonstrates. Another view involves what we might call self -demonstration, whereby an instance of Q comprises the very phenom-enal concept C whose deployment functions as a refl exive device, referring directly to itself. Th e former kind of demonstration is the sort of view Tye and McLaughlin have in mind, while the latter kind is what Block, Balog, and Papineau have in mind on their “quotational” accounts of phenomenal concepts. Each of these views can be combined with one of two broad notions of direct reference. Th e fi rst is a kind of reference or demonstration that serves to reveal substantive content and information about the nature of what is demonstrated. On a theory which holds that subjects are literally acquainted with phenomenal properties, substantive demonstration of Q via the exercise of C is grounded in the intimate ontological and epistemic relation that obtains between the subject and the phenomenal content of her awareness. On the other notion of direct reference, demonstration does not function to reveal information about the nature of a given phenomenal property, but merely acts as a pointer or indexical for the property referred to. In this case, the demonstrating or naming function is most likely imple-mented through some sort of distinctive causal-functional role. Th e particu-lar role of a phenomenal concept state C that directly designates or demonstrates Q will have a content along the lines of “caused by Q” or “this/that state Q.” But since C contains no substantive information about Q (i.e., the subject does not access anything about the nature of Q via her exercise of C), “Q” acts as a mere label for whatever state C (rigidly) desig-nates or demonstrates.

Given their insistence on (iv), it is clearly this latter notion of direct refer-ence and demonstration that is in play for Tye and McLaughlin. But if that is what direct reference (and ultimately direct phenomenal access) comes down to, then their view cannot explain the phenomenally rich access we have to the qualitative character of experience. And as a result, the view of phenomenal access under consideration has no principled means of account-ing for categorical diff erences in phenomenal character and content. Suppose


I look at two large colored squares painted on a wall, a blue square above a red square. (Suppose also that my color experiences are correlated in the normal way with diff erent underlying physical properties.) I fi rst introspec-tively attend to the blueness of the blue square, and then to the redness of the red square, and in doing so, I exercise phenomenal concepts that directly demonstrate the phenomenal qualities I grasp. (Call the properties B and R, and the phenomenal concepts that designate them B* and R*, respectively.) My introspective awareness of the two squares clearly seems to aff ord me a qualitatively diff erent grasp of their respective colors, as my two visual expe-riences diff er in a categorical way with respect to their phenomenal character; in one case I am phenomenally aware of B, in the other case of R. Here is the problem: when I introspectively attend to the color qualities of the squares, if there is no minimally descriptive or substantive content that I grasp via my exercise of B* and R*—no determinate conception or grasp that refl ects the way the phenomenal color properties actually are, but only demonstration—then how am I to account for the qualitative diff erences in the way the squares look to me, that is, the diff erences between B and R? Th e diff erences in the way the squares look to me appear to be categorical, qualitative diff er-ences in the nature of the phenomenal properties presented. But if demon-stration or naming aff ords no positive, determinate grasp of the way the qualities B and R are in themselves, then there exists no explanatory ground for the substantive diff erence in phenomenal content I notice when I deploy B* and R*. 25

One might attempt to dispute the above point by claiming that judgments of phenomenal diff erence can be reliably triggered in other ways, e.g. by the appropriate physical-functional states underlying or realizing B* and R*. Th ere is some experimental evidence indicating that, say, accurate visual dis-criminations need not be grounded in phenomenal discriminations. 26 Discriminating the blue square from the red square need not involve grasping

25 It won’t work to locate the phenomenal diff erences in the distinct conceptual roles of B* and R*, for the original tension among (i)–(iii) re-emerges as a problem about how phenomenal character can be plausibly identifi ed with these functional roles. I am directly and immediately aware of the qualitative character of my experience in a rudi-mentary yet substantive way that precludes its being merely a functional property of the sort suggested here. Moreover, conceptual roles seem by themselves phenomenally blank and therefore cannot constitutively ground the intrinsic, qualitative character of experi-ence. B* and R* have distinct conceptual roles, to be sure, but the diff erences in qualita-tive character are either grounded in the diff erences in intrinsic natures of the properties that are their realizers (quotational account) or else grounded in the diff erences in natures of the properties they designate (recognitional account).

26 See for example, the work of Milner and Goodale ( 1995 ) .


any categorical diff erences in the essences of phenomenal blue and phenomenal red. Yet our judgments about, say, color patches typically do seem to be grounded in this way, by noticing categorical phenomenal diff erences. Th at is, at least, often the way things phenomenally seem to us (and our argument has been broadly introspective from the start). And that observation is compatible with evidence that shows such judgments need not be, or are not always, so triggered. Materialist accounts of phenomenal concepts, like that of Tye and McLaughlin, imply that such judgments are never grounded in categorical phenomenal diff erences.

On the recognitional-demonstrative accounts of Tye and McLaughlin, and given their insistence on (iv), from the standpoint of the subject and the phenomenal concept being deployed, it does not matter whether it is B or some other property R that is the one being demonstrated; a phenomenal concept’s having the form “this state R” instead of “this state B,” is a seman-tic distinction that makes no phenomenal diff erence in my grasp of the two properties. For it is a mere formal diff erence in label or naming, one that does not refl ect any diff erence in the substantive content or nature of B and R. Intuitively, the diff erence between exercising a phenomenal concept that designates or demonstrates B and one that designates R is a diff erence in the natures of the particular phenomenal qualities demonstrated by the con-cepts or tokened in the concepts. Th at is, what marks the diff erence in my phenomenal grasp of B and R is the way those qualities are intrinsically, in their essence. Unfortunately, this explanation is unavailable to Tye and McLaughlin and other proponents of materialist accounts of phenomenal concepts who endorse (iv). 27 Hence we conclude that this type of account lacks the resources altogether to explain the manifest qualitative diff erences that exist between the what-it’s-like or phenomenal feel that accompanies one’s grasping type-distinct phenomenal properties. Pushing the implica-tions of (iv) leads to implausible denial of the phenomenological evidence by failing to acknowledge the intrinsic simplicity of phenomenal character. It allows only a watered-down sense in which phenomenal properties “directly present” themselves, a very attenuated understanding of (i); and this in turn seems to imply that the character of phenomenal awareness is deeply illusory, contrary to the avowals of physicalists who aim to provide a phenomenally realistic account.

27 Th e problem described in this section equally confronts (materialist) “quotational” or self-referential accounts of phenomenal concepts that endorse (iv). While the func-tional mechanism involves self-reference or self-demonstration, this mechanism does not confer information about the nature of the properties in question; it simply functions as a sort of indexical of the form “this state I am in now (B).”


In applying a recognitional-demonstrative account of phenomenal con-cepts to the problem of direct awareness, physicalists such as Tye and McLaughlin carve out a fi nely nuanced position that affi rms and walks a thin line between the following two claims:

i. S directly accesses Q as it truly is in itself. iv. S does not directly access Q’s nature or essence.

Th e natures of the properties we are directly aware of are completely hidden from our awareness, not presented to us in any way at all. We have argued that physicalists cannot plausibly hold both (i) and (iv) together. (i) does justice to the phenomenal data only insofar as it allows a limited revelation thesis (R) and hence excludes (iv). Drawing out the implications of (iv) leads to an implausible denial of the phenomenological evidence and has the consequence that much of the content of phenomenal experience is illusory.

On the other hand, a primitivist theory of phenomenal character—one that proposes ontologically simple phenomenal qualities as the immediate objects of awareness, along with an ontologically primitive acquaintance relation to ground our access to these qualities—eff ectively dispels the ten-sion that arises for physicalist theories. Th is is because primitivism rejects a tenet that proves to be the needless source of the tension, namely:

ii. Q is a physical-functional property.

Th e phenomenal feel of sensory awareness presents us with ontologically simple, non-structural qualities with a determinate intrinsic character. Th ese are “non-physical-functional” in the sense that they cannot be characterized in terms of or reduced to basic, underlying physical properties counte-nanced by a mature physics. A theory is needed that incorporates this fea-ture of the intrinsic simplicity of phenomenal character ( IS ). Primitivism eff ectively integrates central claims about the nature of phenomenal access—e.g., that we grasp phenomenal properties as they really are, and that we have a substantive conception of their character, albeit one that amounts to much less than a sophisticated theoretical description. Primitivism holds that in phenomenal and introspective awareness we come to access phe-nomenal properties as they truly are, under no distinct modes of presenta-tion, and via a substantive (and perhaps minimally descriptive) partial grasp of their natures. On this view, to say that a phenomenal property is pre-sented to a subject as it genuinely is means that the nature of the property, at least in part, is revealed through an intimate acquaintance the subject has with one or more of its instances, in a way that rules out widespread illusion with respect to the way things seem phenomenally to the subject. Such a


view accords well with what we fi nd phenomenal experience to be like, with the manner in which the qualitative character of experience is immediately present to our awareness. Th ere seems to be no other plausible explanation for what phenomenal character is , if it is not the essence of whatever it is that is revealed to us in q -awareness.

Returning to our argument from direct awareness, we discover that the very feature of intrinsic simplicity provides a basis for a direct argument for primitivism about phenomenal character. We can construct this argument by modifying the earlier failed argument of section II:

1. S directly accesses Q as it truly is in itself. 2. S directly accesses Q as a primitive, non-structural property. 3. If S directly accesses Q as it is in itself and S directly accesses Q as

a property of type F, then Q is a property of type F. 4. Therefore, Q is a primitive, non-structural property.

Th e controversial premise—the one that will be contested by Tye, McLaughlin, and other like-minded physicalists—is (2). Th e basis of (2), as we have tried to argue, is simply what is given in perceptual phenomenol-ogy: we have an unmediated awareness of phenomenal character, an aware-ness that does not admit of an appearance/reality distinction with respect to the way things are phenomenally, and the content of that awareness is pre-sented to us as various unanalyzable, non-structural features having a deter-minate qualitative character. As such, they are ontologically distinct from the neuro-physical-functional states of the brain. Th e result is a robust dual-ism of properties for which we seek an explanation.

V. THE DUALIST ARGUMENT FROM CONSCIOUSNESS RECAST AS A FINE TUNING ARGUMENT

To this point, we have been concerned to establish the central assumption underlying the argument from consciousness: the thesis that unfolding con-scious experience involves systematic correlations of non-identical neuro-physical and phenomenal states. Let us call this premise the Dual Property Correlation Th esis (DPC). We have examined the most sophisticated recent materialist challenge to DPC and argued that it fails. We now turn to con-sider the argument from conscious experience as a whole.

Swinburne ( 2004 ) and Adams ( 1987 ) both argue that there can be no—or, more cautiously, that it is very unlikely there can be a—systematic scientifi c or natural connection between physical properties and experiential qualities


that would explain why they are correlated in the patterns that they are. Since an explanation of some sort is required, and the prospect for a natural-istic explanation is not in the offi ng, a theistic design-style explanation is the most plausible remaining option.

Note that, as stated, DPC is hardly an appropriate candidate for scientifi c explanation. Just restricting our attention to human experiencers, there are a huge variety of actual and possible phenomenal states, on the one hand, and neuronal state confi gurations in relevant portions of the brain, on the other. Present science knows a great deal about neurons, the basic elements of neuro-physiological architecture, and some of the mechanisms of com-munication between them. Much is still unknown about the relevant struc-ture-types that subserve psychological operations, but no one doubts that a systematic theory is there to be had. One persuaded of the ontological irre-ducibility of phenomenal qualities will naturally suppose that a mature sci-entifi c classifi cation of them would similarly discern fundamental properties and relations and specify a corresponding combinatorial algorithm. With such a classifi cation scheme in hand, one could then hope to identify gen-eral laws of neural-phenomenal co-variation that are much simpler than brute enumeration of the billions of one-one correlates.

Swinburne dismisses the possibility of any such simplifying account on the grounds that phenomenal qualities are not quantifi able, which is a cen-tral characteristic of scientifi c laws. 28 However, as we have just suggested, it is plausible that qualia exhibit geometrical relations and admit variations of degree (in similarity and diff erence), intensity, extent (of a visual fi eld), and so on. 29 Th eir essentially private nature may preclude exactitude in measure-ment, as a practical matter, but the argument from consciousness concerns the in-principle availability of a scientifi c explanation of correlations, not the practical feasibility of doing so.

Adams allows that qualia may admit such quantifi able descriptions, but believes that two insuperable obstacles to a scientifi c account of conscious-ness remain. First, the occasional fl irtation with panpsychism aside, it is not plausible that any kind of phenomenal properties correlate with anything other than highly structured neural states, or at least states that bear relevant functional similarities to neural states. An hypothesized law, then, would at most chart relations between basic phenomenal qualities and certain physi-cally complex kinds, rather than similarly basic physical kinds. As Adams sees it, such a law would merely describe rather than explain an important fact needing explanation, viz., that phenomenal qualities are correlated with

28 Swinburne ( 2004 : 204) . 29 Cf. Chalmers ( 1996 : 224–5) .


specifi cally neural states, and not with other physical states, whether simple or complex. 30 Th e second obstacle to a scientifi c explanation of conscious experience he expresses thus:

. . . even if we had, from a purely phenomenal point of view, a single uniquely valid spectrum for each sensory modality, we would still face the mind-boggling problem of fi nding a mathematical relationship between the qualia of the diff erent modali-ties. And without such a relationship, our law of nature will not explain why certain brain states produce phenomenal qualia such as red, yellow, and blue, and others produce qualia such as sweet, sour, and salty. 31

We believe that Adams is correct that these two considerations preclude a certain type of explanation, a type of explanation that we are accustomed to seeing science pursue and often achieve, but we deny that they preclude any form of explanation whatsoever. Th e form of explanation that non-panpsy-chist qualia realism precludes is a reductive and maximally unifi ed explana-tion. Our most general science, comprised of the branches of fundamental physics, has achieved great success toward the goal of a highly unifi ed science of the elementary causal factors structuring the universe. It is an open empir-ical question whether a maximally unifi ed theory—a single, simple equation governing a single physical constituent—might be possible for the behavior of at least relatively disorganized and isolated physical systems. Suppose a theory of this sort is not attainable. We take it that this fact alone would not cry out for a further and inevitably purposive style of explanation. If at bot-tom there were, say, two fundamental particles whose interactions were described by a couple of equations, then we should say that these facts pro-vide the deepest explanations of the universe’s dynamics. Th ese facts are not themselves explainable, but even that needn’t, if we are causal realists, be taken to indicate a kind of brutely contingent behavior. We may say that the equations encode the dispositional profi les of the particle types, profi les that are essential (de re, not merely de dicto) to those particles. Of course, it will be scientifi cally unexplained why the universe is ultimately constituted by these sorts of particles, rather than some other sorts, or none at all. But this fact is not germane to the present discussion, as Adams and Swinburne do not intend the argument from consciousness to be just a restricted version of the cosmological argument from contingency, à la Leibniz. 32

Admitting primitive phenomenal qualities that causally interact with cer-tain kinds of structured physical states is also to give up the aspiration for

30 Adams ( 1987 : 255) . 31 Ibid. (257) . 32 One of us has addressed this question elsewhere. See O’Connor ( 2008 ) .


maximal theoretical unifi cation. It means that there are properties that are emergent in a robustly ontological sense, with correspondingly irreducible laws that chart their patterns of instantiation and their contributions to the dynamical evolution of physical systems. But emergence needn’t be under-stood as an objectionably brute aff air of primitive properties occurring without any causal antecedents. If we are causal realists, we shall say that fundamental properties have two distinct types of causal dispositions: the familiar “locally determinative” dispositions that are manifested in even small-scale systems (such as the disposition of negatively charged particles to repel one another, absent countervailing forces) and a disposition in the right kind of organized system to contribute to the generation of an emer-gent property. Th at fundamental physical properties confer such emergent dispositions is a surprising “hidden” fact from the standpoint of one who studies the behavior of fundamental particles in the context of isolated, small-scale systems, as particle physicists do. But it is really no more myste-rious than whatever turns out to be the most fundamental locally determi-native dispositions of matter. Both kinds of dispositions are discovered empirically by successful application of theories that posit them. Neither are transparently “intelligible” to human inquirers in the way Leibniz and Spinoza imagined. 33 If we ask why negatively charged particles are disposed to repel each other, there will be the true, even if not reductively illuminat-ing, answer: the disposition to such interactions is the nature of the prop-erty of negative charge. 34 Likewise, if we ask why neural state N1 gives rise to an experience of phenomenal blue, rather than phenomenal yellow, there will be a true answer involving a fundamental disposition of the fundamen-tal constituent properties of N1 toward just such an eff ect in just such a context, a disposition that is essential to them.

Do the foregoing considerations indicate that the argument from con-sciousness is altogether worthless? In our view, such a negative verdict would be too hasty. Th ere is an explanandum suited to design-style hypotheses in the neighborhood, it’s just that Swinburne and Adams haven’t characterized it properly. We have suggested that the phenomenal realist may reasonably suppose the existence of basic, general laws connecting neural-state types and families of phenomenal-state types (corresponding more or less directly

33 See Robert Adams’s discussion of Lebniz’s claim that an “eff ect is conceived through its cause” ( 1994 : 152ff .) .

34 We needn’t here enter into the dispute between pure dispositionalists, who hold that properties just are clusters of dispositions, and others who hold that dispositions are one aspect of properties alongside another, a certain qualitativity that is distinct for each distinct property.


to distinct sensory modalities). Such laws will encode in part facts about specifi c emergent dispositions of fundamental physical particulars. Here is where we see the potential for design-style reasoning. It seems plausible that there are a variety of ways things might have been with respect to the fun-damental constituents of the world. We do not have in mind the Humean claim that the very particulars there are might have interacted in fundamen-tally diff erent ways. We mean, rather, that there might have been ever so many diff erent sorts of entities having diff erent sorts of basic dispositions from the ones that are manifested in our world. And in particular, it seems a priori rather unlikely that fundamental physical entities should have emer-gent dispositions toward phenomenal qualities. (Th at this is a plausible claim is suggested by the fact that many rash but otherwise reasonable phi-losophers judge the emergentist view to be an utterly implausible hypothesis about our own world, and some are tempted to declare it outright impos-sible.) Yet, given theism, it seems more to be expected, since we may reason-ably suppose the conditional probability of there being agents capable of the kind of experiential life that we enjoy on the hypothesis of theism to be at least not very low, since it is reasonable to think that one of the goods a purposive world designer would wish to see in its creation are creatures of just that sort. 35

Taken this way, the facts underlying DPC are plausibly suited to contrib-ute to the evidence base for the modern fi ne-tuning argument for design. 36 Th is argument is rooted in a vast range of information from cosmology and fundamental physics, appearing since the mid-1970s. Physicists have docu-mented dozens of fundamental respects in which our universe, according to

35 An anonymous referee objects as follows: even if we grant that the probability that there are conscious agents given theism is relatively high, there is no reason to think that conscious experience has to include primitive, non-structured phenomenal properties. So, there is no reason to think that the probability that there would be primitive, non-structured phenomenal properties given theism is signifi cant. We make two comments in reply: (1) Granted, we have not argued above that the account of conscious experience we give is necessarily true of any phenomena properly termed “conscious experience.” But it seems to us plausible to think that it is necessarily true, if true at all, given that our con-cept of conscious experience is closely tied to general features of our own experience. It seems even more plausible that conscious experience of this sort, assuming it isn’t the only sort, is especially valuable, and so something that a divine being would choose to instanti-ate. (2) But we needn’t in any case commit ourselves to either of these claims. For the conditional probability that is at issue doesn’t primarily concern the likelihood of there being primitive vs. structured higher-level properties. Rather, it is that, given that con-scious properties in our world are primitive, the likelihood that such properties would arise from the lower-level structured neuro-physical base properties—i.e. that such psy-cho-physical correlations would hold—is higher on theism than the denial of theism.

36 Several of the paragraphs to follow are adapted from O’Connor ( 2008 ) .


present theory, is exquisitely “fi ne-tuned” for the eventual appearance of biological (and so intelligent and sentient biological) life. What is meant by the neutral term “fi ne-tuned” is that there are apparently contingent fea-tures of the universe, involving such things as ratios of basic particles or forces or the specifi c value of very large numerical constants in dynamical laws, such that had any one of them diff ered appreciably, the universe would not have evolved in a way consistent with the appearance of biological life. Th e details have been documented in several places. 37

It is currently estimated that there are more than thirty examples of fi ne-tuned features of the universe, many of which appear to be independent of the others. Current physics, of course, is rife with speculation about struc-tures more fundamental than those that current theory treats as basic. Informed opinion appears to be that some of the seemingly arbitrary con-stants will be explainable in more fundamental terms, but it is quite unlikely that all will be. It is noteworthy that some current speculative models which promise to explain away some fi ne-tuned features introduce others (as is the case with both chaotic infl ationary theory, which has received signifi cant empirical support, and the at present purely speculative string theory, which contemplates a great simplication of the basic ontology of physics). 38 We will here presume that this opinion is justifi ed.

What, then, are we to make of all this? According to the defender of the fi ne-tuning argument, we should infer that our universe is fi ne-tuned because it was created by an intelligent, powerful designer, who created the right elements in just the right initial conditions as to allow for the eventual emergence of intelligent and sentient life. It is not improbable, and perhaps is more probable than not, that such a being would desire a universe con-taining some form of intelligent life. Th e alternative hypothesis is that the one universe there happens to be exists as a brutely contingent fact. But this is overwhelmingly improbable . Th ink of all the variations on the above con-stants as specifying a space of possibility. Th ose possibilities that are life-sustaining occupy an incredibly tiny portion of this space. Th at the one realized possibility happens to fall into this tiny, yet highly signifi cant por-tion calls out for explanation. Assuming that the existence of such an intel-ligent designer is not as improbable a priori (prior to the consideration of fi ne tuning) as the existence of the universe itself on the chance hypothesis, the most plausible explanation is in terms of intelligent design.

37 Brandon Carter ( 1974 ) fi rst called attention to this data. For a comprehensive treat-ment, see J. D. Barrow and F. J. Tippler ( 1986 ) .

38 Robin Collins develops this point in “Th e Many-Worlds Hypothesis as an Explanation of Cosmic Fine-Tuning: An Alternative to Design?,” currently unpublished.


A version of the argument may be expressed in the formal apparatus of Bayesian confi rmation theory as follows. 39 Let:

L = a universe’s basic parameters must be fi nely tuned for intelligent life to exist in it. E = Our universe permits intelligent life. D = There is a supernatural universe designer.

Where “P(A/B)” is to be read as “the probability of A, given that B is true,” the argument’s proponent contends that:

P(E/L & ~D) is very low. P(E/L & D) is signifi cant. P(D/L) is signifi cantly greater than P(E/L &~D).

Th at is, what we observe (E) is much more likely on the assumption that there is a supernatural designer (D) (given that L is in fact true) than on the denial of that assumption. Bayes’s theorem tells us that this should lead us to upgrade the probability we assign to D (and correspondingly down-grade the probability of ~D) proportionate to the strength of the last two conditional probabilities and the weakness of the fi rst. When a bit of evi-dence is highly unexpected on one theory but not so unexpected on another, the second is to be judged more likely, provided its prior probabil-ity is not so low as to swamp the signifi cance of these conditional probabil-ity diff erences.

Now, we will not here try to assess the overall cogency of this argument, which has been the subject of a growing body of literature. 40 We simply sug-gest that the argument from consciousness is best developed as adding to the data of fi ne-tuning. Given the role that comparative probability judg-ments play in the fi ne-tuning argument, it is doubtful that the a priori probability of the causal properties of the universe’s fundamental constitu-ents being such as to ground physical-phenomenal correlations is suffi -ciently low in itself to signifi cantly bolster the case for a designer. Yet, if our argument in sections III–IV above are sound, the data that consciousness provides are distinctive elements of that case insofar as they are not subject to being explained away through future development of more comprehen-sive physical theories. Hence, the Dual Property Correlation Th esis is best

39 Here we closely follow Neil Manson’s presentation in the introduction to Manson ( 2002 : 7) .

40 In addition to the essays in Manson ( 2002 ) , see Colyvan, Garfi eld, and Priest ( 2005 ) , Collins ( 2005 ) , and Pruss ( 2005 ) . Finally, one of us assesses the argument in O’Connor ( 2008 : chs. 4–5) .


taken as one important and comparatively secure ingredient in an overall cumulative case for theistic design.

REFERENCES

Adams, Robert (1987), Th e Virtue of Faith and Other Essays in Philosophical Th eology (New York: Oxford University Press).

—— (1994), Leibniz: Determinist, Th eist, and Idealist (New York: Oxford University Press).

Balog, Katalin (unpublished ms.), “Th e ‘Quotational’ Account of Phenomenal Concepts.”

Barrow, John D. and Tippler, Frank J. (1986), Th e Anthropic Cosmological Principle (Oxford: Th e Clarendon Press).

Block, Ned (2006), “Max Black’s Objection to Mind-Body Identity,” in Torin Alter and Sven Walter (eds.), Phenomenal Concepts and Phenomenal Knowledge (Oxford: Oxford University Press).

—— and Stalnaker, Robert (1999), “Conceptual Analysis, Dualism, and the Explanatory Gap,” Philosophical Review 108(1): 1–46.

Carter, Brandon (1974), “Large Number Coincidences and the Anthropic Principle in Cosmology,” in M. S. Longair (ed.), Confrontation of Cosmological Th eories with Observational Data (Dordrecht: D. Reidel), 291–8.

Chalmers, David (1996), Th e Conscious Mind (Oxford: Oxford University Press). Collins, Robin (2005), “Fine-Tuning Arguments and the Problem of the Comparison

Range,” Philosophia Christi 7(2): 385–404. —— (unpublished ms.), “Th e Many-Worlds Hypothesis as an Explanation of

Cosmic Fine-Tuning: An Alternative to Design?” Colyvan, Mark, Garfi eld, Jay, and Priest, Graham (2005), “Problems with the

Argument from Fine-Tuning,” Synthese 142: 325–38. Dretske, Fred (1995), Naturalizing the Mind (Cambridge, MA: Th e MIT Press). Gertler, Brie (2002), “Explanatory Reduction, Conceptual Analysis, and

Conceivability Arguments About the Mind,” Noûs 36(1): 22–49. Gillett, Carl and Loewer, Barry (2001) (eds.), Physicalism and its Discontents (Cambridge:

Cambridge University Press). Horgan, Terry and Tienson, John (2001), “Deconstructing New Wave Materialism,”

in Gillett and Loewer (2001), 307–18. —— (2002), “Th e Intentionality of Phenomenology and the Phenomenology of

Intentionality,” in D. Chalmers (ed.), Philosophy of Mind: Classical and Contemporary Readings (Oxford: Oxford University Press), 520–33.

Jackson, Frank (1982), “Epiphenomenal Qualia,” Philosophical Quarterly 32: 127–36. Johnston, Mark (1992), “How to Speak of Colors,” Philosophical Studies 68:

221–64.


Kimble, Kevin (2006), Th e Intentional Structure of Phenomenal Awareness (doctoral dissertation, Indiana University).

Kripke, Saul (1980), Naming and Necessity , 2nd edn. (Cambridge, MA: Harvard University Press).

Loar, Brian (1997), “Phenomenal States,” in Ned Block, Owen Flanagan, and Guven Guzeldere (eds.), Th e Nature of Consciousness: Philosophical Debates (Cambridge, MA: MIT Press).

Locke, John (1690), An Essay Concerning Human Understanding (London). Lockwood, Michael (1993), “Th e Grain Problem,” in Howard Robinson (ed.),

Objections to Physicalism (Oxford: Clarendon Press, 271–91). Manson, Neil (ed.) (2002), God and Design (London: Routledge). McLaughlin, Brian (2001), “In Defense of New Wave Materialism: A Response to

Horgan and Tienson,” in Gillett and Loewer (2001). —— (2003), “Color, Consciousness, and Color Consciousness,” in Quentin Smith

and Aleksandar Jokic (eds.), Consciousness: New Philosophical Perspectives (Oxford: Oxford University Press).

Milner, D. and Goodale, M. (1995), Th e Visual Brain in Action (New York: Oxford University Press).

O’Connor, Timothy (2000), Persons and Causes: Th e Metaphysics of Free Will (New York: Oxford University Press).

—— (2008), Th eism and Ultimate Explanation: Th e Necessary Shape of Contingency (Oxford: Blackwell).

Papineau, David (2002), Th inking About Consciousness (Oxford: Oxford University Press).

Pruss, Alexander (2005), “Fine- and Coarse-Tuning, Normalizability, and Probabilistic Reasoning,” Philosophia Christi 7(2): 405–24.

Smart, “Jack” J. C. (1959), “Sensations and Brain Processes,” Philosophical Review 68: 141–56.

Swinburne, Richard [1979] (2004), Th e Existence of God , 2nd edn. (Oxford: Oxford University Press).

—— (1996), Is Th ere a God? (New York: Oxford University Press). Tye, Michael (1995), Ten Problems of Consciousness: A Representational Th eory of the

Phenomenal Mind (Cambridge, MA: Th e MIT Press). —— (2000), Consciousness, Color, and Content (Cambridge, MA: Th e MIT Press). —— (2002), “Phenomenal Character and Color,” Societa Italiana Filosofi a Analitica,

E-Symposium on Color, Consciousness, and Content <http://host.uniroma3.it/pro-getti/kant/fi eld/tyesymp.htm>.

—— (2003), “A Th eory of Phenomenal Concepts,” in A. O’Hear (ed.), Minds and Persons (Cambridge: Cambridge University Press).

White, Stephen L. (2010), “Th e Property Dualism Argument,” in Robert C. Koons and George Bealer (eds.), Th e Waning of Materialism (Oxford: Oxford University Press), 89–114.

Th e Argument from Consciousness Revisitedscotus/files/O'ConnorKimbleOSPRv.III.pdf7 Th e Argument...

Documents

Transcript of Th e Argument from Consciousness Revisitedscotus/files/O'ConnorKimbleOSPRv.III.pdf7 Th e Argument...